Multi-Task Deep Learning Model for Classification of Dental Implant Brand and Treatment Stage Using Dental Panoramic Radiograph Images

It is necessary to accurately identify dental implant brands and the stage of treatment to ensure efficient care. Thus, the purpose of this study was to use multi-task deep learning to investigate a classifier that categorizes implant brands and treatment stages from dental panoramic radiographic images. For objective labeling, 9767 dental implant images of 12 implant brands and treatment stages were obtained from the digital panoramic radiographs of patients who underwent procedures at Kagawa Prefectural Central Hospital, Japan, between 2005 and 2020. Five deep convolutional neural network (CNN) models (ResNet18, 34, 50, 101 and 152) were evaluated. The accuracy, precision, recall, specificity, F1 score, and area under the curve score were calculated for each CNN. We also compared the multi-task and single-task accuracies of brand classification and implant treatment stage classification. Our analysis revealed that the larger the number of parameters and the deeper the network, the better the performance for both classifications. Multi-tasking significantly improved brand classification on all performance indicators, except recall, and significantly improved all metrics in treatment phase classification. Using CNNs conferred high validity in the classification of dental implant brands and treatment stages. Furthermore, multi-task learning facilitated analysis accuracy.

Secular Decreasing Trend in Plasma Eicosapentaenoic and Docosahexaenoic Acids among Patients with Acute Coronary Syndrome from 2011 to 2019: A Single Center Descriptive Study

Despite intensive lipid-lowering interventions, patients treated with statins develop atherosclerotic cardiovascular disease (ASCVD), and these patients have an increased risk of developing recurrent cardiovascular events during follow-up. Therefore, there is a need to focus on the residual risks in patients in statin therapy to further reduce ASCVD. The aim of this study was to retrospectively investigate the 10-year trend (2011–2019) regarding changes in polyunsaturated fatty acids (PUFAs) in patients with acute coronary syndrome (ACS) in a single center. We included 686 men and 203 women with ACS admitted to Kagawa Prefectural Central Hospital. Plasma PUFAs, including eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), arachidonic acid (AA), and dihomo-γ-linolenic acid (DGLA), were measured at admission for suspected ACS. A secular decreasing trend in the levels of EPA and DHA and the EPA/AA ratio, but not of AA and DGLA, was observed. The analyses based on age (>70 or <70 years) and sex showed that the decreasing trend in the levels of EPA and DHA did not depend on age and remained significant only in men. Further studies are needed to obtain robust evidence to justify that the administration of n-3 PUFA contributes to the secondary prevention of ACS.

Deep Neural Networks for Dental Implant System Classification

In this study, we used panoramic X-ray images to classify and clarify the accuracy of different dental implant brands via deep convolutional neural networks (CNNs) with transfer-learning strategies. For objective labeling, 8859 implant images of 11 implant systems were used from digital panoramic radiographs obtained from patients who underwent dental implant treatment at Kagawa Prefectural Central Hospital, Japan, between 2005 and 2019. Five deep CNN models (specifically, a basic CNN with three convolutional layers, VGG16 and VGG19 transfer-learning models, and finely tuned VGG16 and VGG19) were evaluated for implant classification. Among the five models, the finely tuned VGG16 model exhibited the highest implant classification performance. The finely tuned VGG19 was second best, followed by the normal transfer-learning VGG16. We confirmed that the finely tuned VGG16 and VGG19 CNNs could accurately classify dental implant systems from 11 types of panoramic X-ray images.

Seroprevalence of Antibodies to SARS-CoV-2 in Healthcare Workers in Non-epidemic Region: A Hospital Report in Iwate Prefecture, Japan

BackgroundAs of June 18, 2020, Iwate is the only one of 47 prefectures in Japan with no confirmed coronavirus disease 2019 (COVID-19) cases. Serological survey for COVID-19 antibodies is crucial in area with low prevalence as well as epidemic area when addressing health and social issues caused by COVID-19.MethodsSerum samples from healthcare workers (n = 1,000, mean 40 ± 11 years) of Iwate Prefectural Central Hospital, Iwate, Japan were tested for the prevalence of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) antibodies. Two laboratory-based quantitative tests (Abbott’s and Roche’s immunoassays) and one point-of-care (POC) qualitative test performed simultaneously. Sensitivity and specificity were 100%, 99.6% in Abbott’s immunoassay; 100%, 99.8% in Roche’s immunoassay; 97.8%, 94.6% in Alfa POC test, respectively.ResultsThe laboratory-based quantitative tests showed positive in 4 of 1,000 samples (0.4%) (95% CI: 0.01 to 0.79): 4/1,000 (0.4%) (95% CI: 0.01 to 0.79) in Abbott; 0/1,000 (0%) in Roche. Positive samples were not detected for both Abbott’s and Roche’s immunoassays. The POC qualitative test showed positive in 33 of 1,000 samples (3.3%) (95% CI: 2.19 to 4.41). There were no samples with simultaneous positive reaction for two quantitative tests and a POC test.ConclusionsInfected COVID-2 cases were not confirmed by a retrospective serological study in healthcare workers of our hospital. The POC qualitative tests with lower specificity have the potential for higher false positive reactions than the laboratory-based quantitative tests in areas with very low prevalence of COVID-19.

Association between total dose of ritodrine hydrochloride and pulmonary oedema in twin pregnancy: a retrospective cohort study in Japan

ObjectivePulmonary oedema is recognised as a severe side effect of ritodrine hydrochloride. Recently, the number of twin pregnancies has been increasing. Few studies have reported the association between total dose of ritodrine hydrochloride prior to delivery and pulmonary oedema in twin pregnancy. We aimed to examine this association and determine the optimal cut-off threshold of total ritodrine hydrochloride dose to predict the incidence of pulmonary oedema in twin pregnancy based on obstetric records.DesignRetrospective cohort study.SettingYamanashi Prefectural Central Hospital, Japan.ParticipantsTwo hundred and twenty-six women with twin pregnancy who delivered at Yamanashi Prefectural Central Hospital between September 2009 and November 2016.MethodsThe obstetric records of the participants were analysed. We defined 1 unit of ritodrine hydrochloride as 72 mg per 24 hours continuous transfusion at 50 µg/min to calculate the dose of ritodrine used for tocolysis.Outcome measuresMultivariable logistic regression analysis was performed to examine the association between total dose of ritodrine hydrochloride used for threatened preterm labour and pulmonary oedema, while controlling for potential confounding factors. Then, a receiver–operating characteristic curve was used to determine the optimal cut-off of total ritodrine dose to predict pulmonary oedema incidence.ResultsMean maternal age was 32 (range, 18–46) years; 143 participants were nulliparous (63.3%), 109 had (48.2%) term deliveries and 194 (85.8%) had caesarean deliveries. The overall incidence of pulmonary oedema was 13.7% (31/226). Multivariable analysis showed that the total dose of ritodrine was significantly associated with pulmonary oedema (adjusted OR 1.02; 95% CI 1.004 to 1.03). The best cut-off point to predict the incidence of pulmonary oedema was 26 units (1872 mg) (sensitivity, 61.3%; specificity, 87.8%).ConclusionOur results suggest that consideration of the total dose of ritodrine hydrochloride is helpful in the management of patients with threatened preterm labour in twin pregnancy.